Liking Product Landscape

Liking Product Landscape (LPL) can be easily used for understanding the comparison of products based on consumers' perceptions evaluations.

We are preparing a journal paper where LPL methodology can be described.

Quick Start

Preparation for using LPL methodology

• Download and install anaconda distribution
• Know the fundamentals of python

An example of the using of Liking Product Landscape

First, it is needed to import the python libraries.

import numpy as np
import pandas

The LPL.py file in this repository, which contains the source code of LPL methodology, needs to be downloaded and paste in the actual project folder. For using the LPL code in a new code file, we must add the next line:

from LPL import LikingProductLandscape

It is recommended to capture the data in a .csv file, like the one in this repository Experiments/Exp2/data/wine.csv. For reading a .csv file the command pandas.read_csv can be used, specifically for the data in the experiment 2 we can use:

X = pandas.read_csv('Experiments/Exp2/exp2_wine_data.csv').values

An important step is to read the consumers' evaluations from the file. It is recommended to use columns for evaluations and rows for consumers. Then, we can read the overall liking and the attributes evaluations indicating the column numbers. It is important to mention that first column corresponds to number 0. For example, for reading the evaluations of the data file Experiments/Exp2/data/wine.csv we use:

overall_liking = X[:,[10,16,22,28,34]]
sweetness = X[:,[5,11,17,23,29]]
acidity = X[:,[6,12,18,24,30]]
astringency = X[:,[7,13,19,25,31]]
body = X[:,[8,14,20,26,32]]
fruity = X[:,[9,15,21,27,33]]

Then, we need to organize the data that corresponds to overall liking (ol), jar (attributes perception) and oljar (ol+jar).

ol = overall_liking
jar = np.concatenate((sweetness,acidity,astringency,body,fruity),axis=1)
oljar = np.concatenate((ol,jar),axis=1)

The parameters of LikingProductLandscape must be defined. The data that will be used, it could be ol, jar, or oljar. The consumers' map technique, it could be 'MDS', 'IPM' or 'PCA'. The preference map technique, it could be 'Danzart' or 'SVM'. The recommended parameters are the ones showed in the sample line code:

lpl = LikingProductLandscape(oljar,consumers_map='MDS',preference_map='SVM')

Then, we need to specify the product names and send the overall liking evaluations with the command products_overall_liking.

lpl.products_overall_liking(['W1','W2','W3','W4','W5'],overall_liking)

In addition, the attributes names and their evaluations need to be sended using the command attribute, one per one.

lpl.attribute('Sweetness',sweetness)
lpl.attribute('Acidity',acidity)
lpl.attribute('Astringency',astringency)
lpl.attribute('Body',body)
lpl.attribute('Fruity',fruity)

Finally, we need to execute the LPL methodology and specify the folder where the files will be stored.

lpl.execute(filename='Experiments/Exp2/results/')

For running the whole code, the complete script is:

import numpy as np
import pandas
from LPL import LikingProductLandscape

## Read the data
X = pandas.read_csv('Experiments/Exp2/exp2_wine_data.csv').values

# Read the overall liking and attributes perception
overall_liking = X[:,[10,16,22,28,34]]
sweetness = X[:,[5,11,17,23,29]]
acidity = X[:,[6,12,18,24,30]]
astringency = X[:,[7,13,19,25,31]]
body = X[:,[8,14,20,26,32]]
fruity = X[:,[9,15,21,27,33]]

# Liking Product Landscape
ol = overall_liking
jar = np.concatenate((sweetness,acidity,astringency,body,fruity),axis=1)
oljar = np.concatenate((ol,jar),axis=1)

lpl = LikingProductLandscape(oljar,consumers_map='MDS',preference_map='SVM')
lpl.products_overall_liking(['W1','W2','W3','W4','W5'],overall_liking)
lpl.attribute('Sweetness',sweetness)
lpl.attribute('Acidity',acidity)
lpl.attribute('Astringency',astringency)
lpl.attribute('Body',body)
lpl.attribute('Fruity',fruity)
lpl.execute(filename='Experiments/Exp2/results/')